// The most powerfull GPU


The NVIDIA B200 based systems propels the data center into a new era of accelerating computing
and generative AI, integrating NVIDIA Blackwell GPUs with high-speed interconnect toaccelerate AI performance at scale.
As a premier accelerated scale-up x86 platform with up to 15X faster real-time inference performance with 12X lower cost and 12X less energy, DGX B200 is designed for the most demanding AI, data analytics, and high-performance computing (HPC) workloads, making it the perfect solution unleashing the full potencial of the B200 GPU.

// For the highest workload

Powered by the NVIDIA Blackwell Architecture

Enter NVIDIA Blackwell GPU architecture, the world’s largest GPU, built with the specific purpose of handling data center-scale generative AI workflows with up to 25X the energy efficiency of the prior NVIDIA Hopper GPU generation.
This technical brief introduces the benefits of NVIDIA Blackwell in detail, including the next-generation superchip, Grace Blackwell GB200, and the next-generation, highperformance HGX systems, NVIDIA HGX B200 and HGX B100.

Accelerate Next-Generation Workloads

  • Generative AI
  • Large language model (LLM) training and inference
  • HPC simulations

GPU Memory


Single-Precision Performance


AI Tensor Performance (FP8)